Swedish ASIC miner company kncminer.com

[DPoS]

Jupiter is now >= 500GH.

http://www.youtube.com/watch?v=JFOcI8nK2xI&hd=1

497.1 average (22 minutes). Still pretty good

   Hashrate Average   Weighted Shares
3 hours   471.52 Gh/s   1185664
22.5 minutes   467.09 Gh/s   146816
256 seconds   583.90 Gh/s   34803
128 seconds   559.10 Gh/s   16663

[1715158241]

1715158241

[1715158241]

1715158241

[1715158241]

1715158241

[ASIC-K]

Does this mean the Saturn will have improved speeds also? I keep hearing about the Jupiter.

they are all the same chips!

[ImI]

Guys, they are just tweaking with that miner. I wouldnt read to much from that.

They said 576 GH/s and thats all that matters at the moment.

[Loredo]

Reading this thread is probably the most fun I've had since the Mad Dog left Mike on WFAN.  But, for me, at least, the fun's over.  

Again, well done, KnC/OrSoC.  

Here's a song to describe KnC -and bitfury.  Jerry Riopelle, with the fairly hot Tara Austin, singing Walkin' on Water.

https://www.youtube.com/watch?v=Wq7K1yL_enA

[Wesly]

So far we've only seen a prototype...  Bitfury still took a few weeks from the prototype boards to mass production boards, so I wouldn't be surprised if we didn't see en masse shipping for a little while.  It'll be interesting to see how fast KnC gets these out, I'm always dubious until I see people on the forum getting their products.

If they do begin shipping massive quantities immediately, I'm really curious if they'll pull an "Avalon" and crank the price of their devices once they realize demand is outstripping supply.

Mostly I'm happy to see competition -- because the customers will be the ones winning in the end once these companies start trying to murder each other in sales.

"crank the price"?  They had already lowered their prices for November shipment and even that are looking expensive now compare to what Hashfast and Cointerra are offering.  The Nov batch will only ROI if their competition slips on their shipment dates.  Avalon was able to raise their price to absurd level because there is no real competition and was really the only game in town.  Those days are over (thanks goodness).  I hope the KNC miners will finally put nails on Avalon, BFL and ASICMiner's coffins, all 3 are greedy bastards.

[Xian01]

Kudos to the KNC folk. Against some serious community skepticism and trolling, they appear to have pulled it off.

Nice to see more players entering the hardware arena. Looking forward to seeing the price of these guys in December.

Would be amazing to place an order for a product and have it ship within 7 days.

Cheers, gents.

[dwdoc]

Does this mean the Saturn will have improved speeds also? I keep hearing about the Jupiter.

Divide by 4? ~500/4=~125

576/4 = 144/chip

Mercury 144

Saturn 288

Jupiter 576

[frankenmint]

Does this mean the Saturn will have improved speeds also? I keep hearing about the Jupiter.

Divide by 4? ~500/4=~125

576/4 = 144/chip

Mercury 144

Saturn 288

Jupiter 576

How much more step up is possible though?  I think we should accept up to 80 percent of 560

[timmmers]

Couriers all over the world will be reporting being leapt upon by grinning maniacs for the next few weeks. 

[2112]

One last thing to mention, and since the naysayers have to fall back to the "won't NOI" sour grapes position, it's sort of on topic here. 

That is, how bogus all the linear or constant-growth exponential calculators are, and why only a troll or fool would quote, let alone believe, those results.

There's a function used in mathematical biology modeling named the Gompertz function.  It's got a rich literature, and a pretty useful Wiki page. 

Basically, say you have a fixed food supply in a fixed space for a single-cell population.  The population will start to grow, at a faster and faster rate, reaching a maximum growth rate, and then, as food for any individual becomes relatively scarce with the increasing competition from growing population, growth slows and ultimately flatlines until there are changes in the external conditions, or mutations with the population that give a new strain a (temporary) competitive advantage.

If that doesn't sound like bitcoin mining to you, maybe you're in the wrong racket.

Somebody ought to produce a calculator that entails Gompertz parameter estimation (which are functions of power consumption, absolute hardware costs, etc.) to give people a chance to estimate the real economics, instead of calculations which have (I'm exaggerating here) more petahash on line in 2015 than there are gigawatts of electricity to power the network. 

I know, for a fact, that some exist.  Just not yet, it seems, in the public space.   

This wont be of much use, IMHO. If you try to squeeze Bitcoin into the Gompertz model you'll get an equivalent of mutation every two weeks (more precisely at every difficulty readjustment). I'm not saying it would be completely useless, I just don't see much of the predictive power from the formulation that you've described.

[dhenson]

One last thing to mention, and since the naysayers have to fall back to the "won't NOI" sour grapes position, it's sort of on topic here. 

That is, how bogus all the linear or constant-growth exponential calculators are, and why only a troll or fool would quote, let alone believe, those results.

There's a function used in mathematical biology modeling named the Gompertz function.  It's got a rich literature, and a pretty useful Wiki page. 

Basically, say you have a fixed food supply in a fixed space for a single-cell population.  The population will start to grow, at a faster and faster rate, reaching a maximum growth rate, and then, as food for any individual becomes relatively scarce with the increasing competition from growing population, growth slows and ultimately flatlines until there are changes in the external conditions, or mutations with the population that give a new strain a (temporary) competitive advantage.

If that doesn't sound like bitcoin mining to you, maybe you're in the wrong racket.

Somebody ought to produce a calculator that entails Gompertz parameter estimation (which are functions of power consumption, absolute hardware costs, etc.) to give people a chance to estimate the real economics, instead of calculations which have (I'm exaggerating here) more petahash on line in 2015 than there are gigawatts of electricity to power the network. 

I know, for a fact, that some exist.  Just not yet, it seems, in the public space.   

This wont be of much use, IMHO. If you try to squeeze Bitcoin into the Gompertz model you'll get an equivalent of mutation every two weeks (more precisely at every difficulty readjustment). I'm not saying it would be completely useless, I just don't see much of the predictive power from the formulation that you've described.

Anyone else reminded of that scene in goodwill hunting in the bar with that blonde long haired tool?

[nightengale]

KnC have definitely given us their best, and not a moment too soon as BitSyncom postures with news of impending Tradehill auctions. Looking forward to information on KnC shipping.

[Vycid]

One last thing to mention, and since the naysayers have to fall back to the "won't NOI" sour grapes position, it's sort of on topic here.  

That is, how bogus all the linear or constant-growth exponential calculators are, and why only a troll or fool would quote, let alone believe, those results.

There's a function used in mathematical biology modeling named the Gompertz function.  It's got a rich literature, and a pretty useful Wiki page.  

Basically, say you have a fixed food supply in a fixed space for a single-cell population.  The population will start to grow, at a faster and faster rate, reaching a maximum growth rate, and then, as food for any individual becomes relatively scarce with the increasing competition from growing population, growth slows and ultimately flatlines until there are changes in the external conditions, or mutations with the population that give a new strain a (temporary) competitive advantage.

If that doesn't sound like bitcoin mining to you, maybe you're in the wrong racket.

Somebody ought to produce a calculator that entails Gompertz parameter estimation (which are functions of power consumption, absolute hardware costs, etc.) to give people a chance to estimate the real economics, instead of calculations which have (I'm exaggerating here) more petahash on line in 2015 than there are gigawatts of electricity to power the network.  

I know, for a fact, that some exist.  Just not yet, it seems, in the public space.    

This wont be of much use, IMHO. If you try to squeeze Bitcoin into the Gompertz model you'll get an equivalent of mutation every two weeks (more precisely at every difficulty readjustment). I'm not saying it would be completely useless, I just don't see much of the predictive power from the formulation that you've described.

Anyone else reminded of that scene in goodwill hunting in the bar with that blonde long haired tool?

A little.

The two-week period actually means the continuous Gompertz model would be only an approximation of the discontinuous hashrate-to-bitcoin-per-unit-time ratio, which is what we are interested in.

Anyway, the important thing here is that the Gompertz model is a poor estimator for explosive growth, because you typically get overshoot in undamped, underdamped, and overdamped systems. The EQUILIBRIUM trajectory of a system adjusting to a new carrying capacity is the Gompertz function, but we are not in an equilibrium state, and the system is underdamped.

The real growth trajectory is likely to look something like this: http://upload.wikimedia.org/wikipedia/commons/9/9c/Sine_integral.svg

http://en.wikipedia.org/wiki/Overshoot_(signal)
http://en.wikipedia.org/wiki/Overshoot_(population)

[erk]

Does this mean the Saturn will have improved speeds also? I keep hearing about the Jupiter.

Divide by 4? ~500/4=~125

576/4 = 144/chip

Mercury 144

Saturn 288

Jupiter 576

They are useless targets, when 2PH/s of KNCminer gear suddenly hits the net, the machines that make the most profit will be the ones that under clock not over clock. I would be trying to under clock the Jupiter to see if I could get it down to 1watt per GH/s so it can still make profits against an army of Bitfury miners. This more is better nonsense is what's killing BTC mining.

The people that have opted to pay 12 months upfront @ $330/mth. hosting, you can consider that a $0.60c per kWh power cost in your calculations.

[Phoenix1969]

How much more step up is possible though?  I think we should accept up to 80 percent of 560

While people claim 576 I saw < 500gh after a longer period of time. Could be due to the heating issues, but I'd say hope for 500... If they deliver 576, it would be awesome.

I had a jalapeno that would hash for like 10GH for few minutes then drop down to 8Gh.

Just to get YOU up to "speed"

[erk]

Does this mean the Saturn will have improved speeds also? I keep hearing about the Jupiter.

Divide by 4? ~500/4=~125

576/4 = 144/chip

Mercury 144

Saturn 288

Jupiter 576

They are useless targets, when 2PH/s of KNCminer gear suddenly hits the net, the machines that make the most profit will be the ones that under clock not over clock. I would be trying to under clock the Jupiter to see if I could get it down to 1watt per GH/s so it can still make profits against an army of Bitfury miners. This more is better nonsense is what's killing BTC mining.

The people that have opted to pay 12 months upfront @ $330/mth. hosting, you can consider that a $0.60c per kWh power cost in your calculations.

$330/30 days = $33/day
($33/day)/24 hours = $0.45833333c per kWh

Am I missing something here?

Probably a just a bug in the monthly electricity formula http://mining.thegenesisblock.com/ uses. It actually says 0.73c per kWh if you use the default values for a Jupiter instead of punching in the updated specs from earlier today like I did. But it highlights a couple of things, run it as fast as you can if your power is prepaid. Run it as slow as you can before profit starts to slide if your electricity is pay as you go.

[DimensionsOfHell]

The people that have opted to pay 12 months upfront @ $330/mth. hosting, you can consider that a $0.60c per kWh power cost in your calculations.

From what I calculate, I get the following:

$3960 (Hosting cost) - $150 (At least $150 for the PSU you are NOT having to buy) - $150 (approximately what I paid for my Saturn shipping, which you don't have to pay). So...

   $3960
   -$150
   -$150
________
   $3660

(($3660/ 12) [months] / 30 [days])/24 [hours]

$3660/12 = $305 price per month

$305/30 = $ 10.17 per day

$10.17/24 = $ 0.4236 per hour

And that price is without counting cooling cost and internet connection; and I'm sure they would have backup internet, and backup power (I assume, as any decent data center would), which most people do not have.

I personally wouldn't (and didn't) go for hosting, but I can see why someone might. You don't have to worry about anything. Just sit back and watch the satoshi's come in.  

[arousedrhino]

Does this mean the Saturn will have improved speeds also? I keep hearing about the Jupiter.

Divide by 4? ~500/4=~125

576/4 = 144/chip

Mercury 144

Saturn 288

Jupiter 576

They are useless targets, when 2PH/s of KNCminer gear suddenly hits the net, the machines that make the most profit will be the ones that under clock not over clock. I would be trying to under clock the Jupiter to see if I could get it down to 1watt per GH/s so it can still make profits against an army of Bitfury miners. This more is better nonsense is what's killing BTC mining.

The people that have opted to pay 12 months upfront @ $330/mth. hosting, you can consider that a $0.60c per kWh power cost in your calculations.

$330/30 days = $33/day
($33/day)/24 hours = $0.45833333c per kWh

Am I missing something here?

Probably a just a bug in the monthly electricity formula http://mining.thegenesisblock.com/ uses. It actually says 0.73c per kWh if you use the default values for a Jupiter instead of punching in the updated specs from earlier today like I did. But it highlights a couple of things, run it as fast as you can if your power is prepaid. Run it as slow as you can before profit starts to slide if your electricity is pay as you go.

Sorry but that still doesn't make any more sense. On that calculator you input the kWh and that is a fixed variable, I don't see how the specs of the Jupiter could have any effect on the kWh rate. Its your fixed cost for electricity.

[erk]

The people that have opted to pay 12 months upfront @ $330/mth. hosting, you can consider that a $0.60c per kWh power cost in your calculations.

From what I calculate, I get the following:

$3960 (Hosting cost) - $150 (At least $150 for the PSU you are NOT having to buy) - $150 (approximately what I paid for my Saturn shipping, which you don't have to pay). So...

   $3960
   -$150
   -$150
________
   $3660

(($3660/ 12) [months] / 30 [days])/24 [hours]

$3660/12 = $305 price per month

$305/30 = $ 10.17 per day

$10.17/24 = $ 0.4236 per hour

And that price is without counting cooling cost and internet connection; and I'm sure they would have backup internet, and backup power (I assume, as any decent data center would), which most people do not have.

I personally wouldn't (and didn't) go for hosting, but I can see why someone might. You don't have to worry about anything. Just sit back and watch the satoshi's come in.  

And if you say the unit draws about 700watts, it's was originally 640watt, but now it's going faster and I don't know the real figure yet so I am guessing, you get 0.4236 /0.7 = 0.60cents per kWh.

[dhenson]

I'm really hoping this turns out to be true (regarding hosting):

Hashing Bitcoins within a few hours of it being assembled at the factory